Nearest-neighbor mixture models for non-Gaussian spatial processes
From MaRDI portal
Publication:6203346
DOI10.1214/23-BA1405arXiv2107.07736OpenAlexW4388952054MaRDI QIDQ6203346FDOQ6203346
Bruno Sansó, Xiaotian Zheng, Athanasios Kottas
Publication date: 27 February 2024
Published in: Bayesian Analysis (Search for Journal in Brave)
Abstract: We develop a class of nearest-neighbor mixture models that provide direct, computationally efficient, probabilistic modeling for non-Gaussian geospatial data. The class is defined over a directed acyclic graph, which implies conditional independence in representing a multivariate distribution through factorization into a product of univariate conditionals, and is extended to a full spatial process. We model each conditional as a mixture of spatially varying transition kernels, with locally adaptive weights, for each one of a given number of nearest neighbors. The modeling framework emphasizes the description of non-Gaussian dependence at the data level, in contrast with approaches that introduce a spatial process for transformed data, or for functionals of the data probability distribution. Thus, it facilitates efficient, full simulation-based inference. We study model construction and properties analytically through specification of bivariate distributions that define the local transition kernels, providing a general strategy for modeling general types of non-Gaussian data. Regarding computation, the framework lays out a new approach to handling spatial data sets, leveraging a mixture model structure to avoid computational issues that arise from large matrix operations. We illustrate the methodology using synthetic data examples and an analysis of Mediterranean Sea surface temperature observations.
Full work available at URL: https://arxiv.org/abs/2107.07736
Cites Work
- Efficient Algorithms for Bayesian Nearest Neighbor Gaussian Processes
- Dependence Modeling with Copulas
- Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations (with discussion)
- A Latent Gaussian Markov Random-Field Model for Spatiotemporal Rainfall Disaggregation
- Model-Based Geostatistics
- Modeling Flat Stretches, Bursts, and Outliers in Time Series Using Mixture Transition Distribution Models
- Pair Copula Constructions for Multivariate Discrete Data
- Approximating Likelihoods for Large Spatial Data Sets
- Non‐Gaussian geostatistical modeling using (skew) t processes
- A Primer on Copulas for Count Data
- Bayesian Nonparametric Spatial Modeling With Dirichlet Process Mixing
- A general framework for Vecchia approximations of Gaussian processes
- A Bayesian analysis of some nonparametric problems
- Graphical models
- Nonseparable dynamic nearest neighbor Gaussian process models for large spatio-temporal data with an application to particulate matter analysis
- Statistical modeling of spatial extremes
- Conditional specification of statistical models.
- Geostatistical Modelling Using Non-Gaussian Matérn Fields
- A stochastic space-time model for intermittent precipitation occurrences
- A Bayesian prediction using the skew Gaussian distribution.
- Bayesian Prediction of Transformed Gaussian Random Fields
- Compatible Conditional Distributions
- Strength of tail dependence based on conditional tail expectation
- Maximum likelihood estimation of regression parameters with spatially dependent discrete data
- Estimation of Copula Models With Discrete Margins via Bayesian Data Augmentation
- On Construction and Estimation of Stationary Mixture Transition Distribution Models
- Spatial Matérn fields driven by non-Gaussian noise
- Vecchia-Laplace approximations of generalized Gaussian processes for big non-Gaussian spatial data
- Computationally efficient multivariate spatio-temporal models for high-dimensional count-valued data (with discussion)
- Factor Copula Models for Replicated Spatial Data
- Non-Gaussian Bayesian Geostatistical Modeling
- Title not available (Why is that?)
- Modeling skewed spatial data using a convolution of Gaussian and log-Gaussian processes
- Nonparametric Bayesian inference in applications
- Highly Scalable Bayesian Geostatistical Modeling via Meshed Gaussian Processes on Partitioned Domains
- A Space-Time Skew-t Model for Threshold Exceedances
- Bayesian Spectral Modeling for Multiple Time Series
- Sparse Cholesky Factorization by Kullback--Leibler Minimization
- Multivariate Type G Matérn Stochastic Partial Differential Equation Random Fields
- Bayesian Hierarchical Models With Conjugate Full-Conditional Distributions for Dependent Data From the Natural Exponential Family
- Multi-scale shotgun stochastic search for large spatial datasets
This page was built for publication: Nearest-neighbor mixture models for non-Gaussian spatial processes
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6203346)
